powered by Jive Software

Spark client hit high CPU

I have a strange situation where two users of Spark 2.5.8 can successfully communicate with with most other openfire users but when one user tries to open a chat the target users Spark hits high CPU and he has to use task manager to exit the app.

However if the traget user mentioned above open a chat with the same user that caused the issue everything is fine.

As I said this only happens with this two users and the issue is just one way.

We have removed all traces of the client from both machines but the issue still remains…

Any ideas?

Was there ever a solution to this? I have the exact same issue.

You can try 2.6.0 RC1 version http://www.igniterealtime.org/downloads/beta.jsp

I had already tried v2.6.0 RC1 and unfortunately it has the exact same results. Any other suggestions?

I was able to resolve this issue for myself by performing the following:

On the PC that hits a high CPU when contacting a 2nd user, locate the transcript files of the user that peaks your CPU and locks up Spark when you attempt to contact them and delete them.

There should be two transcript files for each user you have had a conversation with and and are located here…

C:\Documents and Settings(user profile name here)\Spark\user\user_account_name_here@server_name_here\transcripts

I closed Spark and reopened it just to be safe. After I logged back on I was able to contact the same user without an issue.

Example transcript file names would be…

johndoe@servername.xml

johndoe@servername_current.xml

Probably those transcripts were huge or contained something which caused Spark to go mad. It is a known thing that huge history files can hog Spark, but rather when trying to open the history, not just contacting the person.

The log file was 140k. I have others that are 400k - 500k and they have not been problematic but who knows. Just thougtht I would mention it.

I know this is an old thread, but I have a single user (Win 7 x64) that keeps having this issue with 2.6.3. Just thought I’d mention the path to these files in Win 7…

C:\Users(user profile)\AppData\Roaming\Spark\user(spark user)\transcripts

Also, rather than delete individual transcripts, I just rename the transcript folder to transcript.old. The transcript folder will then be created anew when the user logs in again and has another conversation. This way, if you ever need to get back into the old trascripts to find something important, they’re still available, just not to Spark directly…